gr5

Cold pull temp for PLA is 90C. ABS is around 130C if I remember right (or is it 140C?). If you can't pull it out at a given temp raise the temp by 10C steps until you can. You want a strong force - roughly 10 pounds or 5 kg.

ABS to PLA is the tricky direction to switch. PLA to ABS is not an issue. You need to clean out all ABS out of the feeder teeth and the bowden tube. Especially if the colors are different. I also strongly recommend you do a cold pull when switching filaments. In fact I do this every time I even just change the color. To do a cold pull (aka atomic method) with ABS set the nozzle temp to >200C (220 is fine) and shove the filament into the nozzle good, then set the temp to 130C, wait for it to get down to 130C and then pull firmly the ABS filament out the feeder. When the filament is half way through the bowden let it cool another 10 seconds before pulling it all the way through the feeder. The resistance from the motor is small but it's best to pull it very fast as the tip goes through the feeder to reduce likelyhood that it breaks off inside the feeder. Then also cut off the tip of the filament if you see any dirt/spots on it. This will keep your nozzle relatively clean.

Awesome Bas. In my tests: http://umforum.ultimaker.com/index.php?/topic/4222-pulling-force-of-um-extruder/?p=34887 The UM2 feeder motor slips somewhere around 4.5 to 5.7kg or (multiplying by 9.81 m/s/s) 44 to 56 Newtons when using the default current (I believe 1.25A limited) although the motor can exert more force when spinning faster before the current limiting feature allows it to slip back.

Yes, bend back the fan. Also after bending it back listen for the limit switch and make sure when it clicks there is still a gap between the fan and the wall. If not the limit switch may have moved or it may be broken. If the limit switch is broken it is obvious because homing the X axis would be incredibly loud. It does the "death rattle" when the limit switch isn't working. Still the switch may have moved and you might want to bend it back by .5mm or so after you bend the fan duct back.

You should only have to put things into the "override table" that are different. For everything else it should take cura settings. So you should be able to leave out "platform adhesion type" if you want it the same for all the objects.

In expert settings under retraction set those min distances to zero and see if that fixes it. You don't have to print anything to see what will happen - after it slices you can look in layer view and retraction is displayed graphically by showing a small vertical blue line attached to the horizontal movement line. That vertical line is not an actual movement but instead represents the act of retraction at the start of the move.

Mine also moved up and down when I first got it and for the first month or so. Then I got a clog and had to take the head apart. When I put it together again I tightened the bowden before tightening the 4 thumb screws. As I tightened the 4 thumb screws the bowden-holder thing got higher and higher until it was all the way up and now I have the little clip on there but it does nothing. So loosen the 4 screws quite a bit (maybe 3mm?) then shove the bowden FURTHER in. Then tighten the 4 screws. You won't need any clips or anything and it will all be solid and not move.

backlash and play are the same thing. There is an article defining these words on wikipedia. That damper definitely looks to be a problem. I recommend you remove it. That looks like around .5mm play right there. So I would expect .5mm error on your Y axis on your circles meaning they would be ellipses that are shorter in the Y direction.

There's lots of possibilities. What speed and layer height are you talking about? One likely possibility - take apart the head and the white teflon piece that touches the nozzle could be deformed - take that out and slide some filament through it to see if it is sliding nicely. If not you might need to drill it out with a 3 or 4mm drill bit. Another possiblity - your 3rd fan stopped working - sometimes the wires fall out of the connectors when the head moves around. Make sure it is working - the fan in the back of the print head (not the side 2 fans). But what layer height/speed are we talking about here?

Dropbox is good - you can put it in the public folder and send me a link for example. If it has tons of polygons you can reduce them with meshlab - here's instructions (meshlabis free): http://www.shapeways.com/tutorials/polygon_reduction_with_meshlab

I knew you would figure that out eventually. You can see the text is backwards in the photo. I used Cura to rotate the text 180 degrees which flips it over for you so you don't have to do it in CAD.

Expect lots more headaches. The easiest thing to do is to take your reel of filament and cut a bit off the tip (100mm?) and then feed the tip through the feeder (you have to push very hard to get it started but once the wheel moves aside it moves easily) until it hits the existing filament in the tube. Then turn on the printer and go to the MOVE FILAMENT menu. Under the ADVANCED menu. Wait for the nozzle to heat up to 180C or hotter and then start spinning the control until you get some filament feeding nicely. At this point you can print something or you can turn that wheel for 10 minutes until you get the broken filament all sent through the nozzle. Alternatively you can do "change filament" and feed in the new filament when it tells you to and then just print. Retraction won't work until you get the broken part passed through the nozzle.

I'm not convinced it's backlash but it might be. Backlash on UM2 is much more rare. Anyway, the feet are smaller for only one reason - shrinkage (possible 2: backlash) but circles are small for 3 reasons and so one expects them to be much smaller than desired. 1) Shrinkage 2) STL doesn't support circles so instead you give it a bunch of lines but for some reason CAD software always puts the lines *inside* the circle and never outside. Don't know why. So the fewer the lines the larger the error. Yet if all your circles are 50 sided polygons that's just silly. 3) Rubber band effect. For PLA it has a long way to cool before it gets close to glass temp (unlike ABS). This starts happening the moment it comes out the nozzle so it is already cooling a bit and shrinking so the PLA comes out kind of like liquid rubber bands - pulling towards the nozzle which pulls in on the part and especially for inside circles, there's nothing to keep the PLA from pulling in but mostly air. On outer perimeters it's not a problem as the part keeps the "rubber band" from slipping in as much.

The right most print doesn't look as good as the leftmost one. The differences between these two parts (first and last) is a similar issue to the photos of the blue part that the original poster posted. The middle two parts don't have the right lighting to decide what's what.

Your Y axis friction looks fine. Your X axis has issus - that ticking is the belt riding up on the pulley I think. You need to check your alignment - the X belts should be directly below the X rods. Directly. As you slide the X to the left then you can check the left end of both X rods and see if the belts are lined up perfectly. At least one of them isn't I think. Loosen that pulley and slide it along the shaft until it's perfectly under the rod then retighten. And tighten the pulley set screw a lot! The hex tool should twist some. If you make another video, pluck the belts with the head in a corner. All 8 top/bottom halves of the 4 long belts and also the 2 short belts (either side). I'm not 100% certain you have backlash. It sure looked like it but maybe it's fine other than the clicking (minor issue). You said you checked the endcaps right? And if you loosen them then the head doesn't move any easier? You can also put a few drops of light oil on all 6 rods (4 thick rods, 2 thin rods) although they shouldn't need it. Any light oil is fine but better to use oil with as few additives as possible (WD40 has additives that you are better off without). But any light oil should be fine.

You should be able to move the head with one finger on each block on the sides. Easily. So that is one finger each hand - each finger pushing on one side block. Try this with the X axis also. If you pluck the long belts like strings they should make a low pitched note. The short belts (to the steppers) should be much tighter - quite tight. Maybe 3Kg force pushing down on the stepper while tightening? In one of the photos from overhead one of the curved edges was almost not touching the second shell layer inside. I think that's the main problem. Higher temperature may also help but I think this is mainly "backlash".

Now that I see your photo I changed my mind! The *walls* aren't bonding. I think you have backlash (also known as "play"). This can be caused by friction too high or belts too loose. Probably belts too loose. Try sliding both X and Y stepper down a bit to tighten the belts. You might have to tighten you long belts also. Paradoxically, if the long belts are too tight then friction will be too high and you can also get the same symptoms. Also if friction is high check that if you loosen some of the 8 endcaps maybe friction will lower drastically.

Does you extrusion look good? Ignoring the first layer, do the lines touch nicely? Or is there a gap between every line as seen from above? If extrusion is good then you need more heat. I only ever print 245C with ABS. That's only 5C hotter than you. I suspect your temperature is off a bit - maybe the probe is too far from the nozzle tip? Maybe the probe is off by 15C? Try raising the temp another 10C or so to about 250C - that's what I would try. But if you get it too hot you can get a clog. When layers separate you usually need more heat.

On the UM2 you don't need any clips to tighten the bowden. Just loosen the 4 long thumb screws about 5mm, then with no clip, push down on bowden and then while still pushing down lift the bowden-holder-ring thing. Then tighten the 4 thumbscrews. Don't tighten the bowden too much or it will be pushing down hard on the white teflon piece (along with the spring). That teflon piece can deform after many weeks of printing at high temps. On another note: I almost always use brim. It's very easy to peel off. I only use skirt for very small parts. Smaller than 25mm across. And regarding Z acceleration - no - it's not zmax. It has the letter A and Z in it. Like "Z Accel" maybe? "ZA"? Not sure. Gotta go out now. Maybe it's not in the UM2 motion menu but I think it is.

I really wish Cura would put a comment in the code for when a bridge starts and ends. Then you could potentially increase flow a bit or at least slow down acceleration a lot (maybe 1000mm/sec/sec) with a plugin.

It's does retract but it keeps extruding (very very slowly) anyway due to heat expansion. You can't stop it. The best fix is to print a skinny tower on either side of the part. This will avoid the "cool head lift" feature you are talking about and keep the tips of the rose from getting too hot. But usually these cool head lift blobs are easy enough to remove/break off.

It may take a week or two to get a resonse from support. Meanwhile - did you probe the PT100 directly? Or at the connector? The most common point of failure is at the connector itself and resoldering that connector can often fix the problem. You can order PT100 parts online. They are very cheap but rarely fail.

I had the same problem - check the wires to the temp sensor (the thinner two wires). Make sure they are screwed down tight. If they are then you might need to reflow (reheat) the solder. You might violate your warranty if you don't let UM Support know about this first - I think the newer UM2's have a sticker on this connector warning you to contact support first.

I haven't experimented much at all with "glues". I just follow other's advice. glue stick or wood glue mixed with water and painted on work incredibly well for PLA (on rare occasions it sticks so well it removes a tiny sliver of glass!). hairspray or abs juice works very well for ABS. I think you can mix and match these products somewhat but not sure. I did some experiments actually with different temperatures and with the glue stick here: http://umforum.ultimaker.com/index.php?/topic/3404-printing-on-glass/ It was very informative - basically there is a sudden temperature change where it sticks incredibly well above a certain temp and horribly below the temp and it doesn't matter how much above or below the threshold you are.

But noone else is having PID issues and lots of people have upgraded the firmware. Daid didn't mention any changes in the PID values or the PID code so I really doubt it's firmware. Plus this is a somewhat common problem. Every time the print head moves back and forth the wires get stressed a little bit. If the wires short a bit that lowers the resistance which indicates a lower temperature.